Post-ablation atrial and ventricular dysfunction on cardiac magnetic resonance and risk of atrial fibrillation recurrence in patients with preserved ejection fraction

Background Catheter ablation (CA) is an established rhythm control strategy for atrial fibrillation (AF), yet AF recurrence remains frequent. Structural and functional abnormalities in the atria and ventricles, including diffuse myocardial disease and inflammation, may sustain a pro-arrhythmic substrate even after apparently successful ablation. Cardiac magnetic resonance (CMR) with parametric mapping and atrial strain offers comprehensive assessment of this post-ablation phenotype. Objectives To characterize atrial and biventricular structure and function on CMR performed at least 3 months after first AF ablation and to identify CMR-derived markers associated with 1-year AF recurrence. Methods Fifty patients with AF (52% male, 68±10 years) undergoing first-time radiofrequency CA and twenty age-matched individuals without AF underwent CMR ≥3 months after the index procedure. Atrial function was assessed using phasic volumes and reservoir, conduit and active strain. Ventricular structure and function were evaluated by biventricular volumetry, left and right ventricular global strain and left ventricular (LV) native T1, T2 (as an index of myocardial inflammation) and extracellular volume (ECV) mapping. Patients were followed for 12 months. AF recurrence was defined as any documented atrial tachyarrhythmia lasting ≥30 seconds after a 3-month blanking period, in line with consensus recommendations. Receiver-operating-characteristic analyses and logistic regression were used to identify predictors of recurrence, and Youden-derived cut-offs informed a dichotomous “high-risk” CMR phenotype for Kaplan–Meier analysis. Results Patients with AF recurrence at 1 year showed more impaired left atrial (LA) active emptying function (14±12% vs 26±14%, P =0.009) and active strain (−4.5±4.3% vs −10.6±6.8%, P =0.004) than those without recurrence. LV native T1 (1040±55 ms vs 985±43 ms, P =0.007), T2 (53.6±1.6 ms vs 50.6±2.5 ms, P =0.002) and ECV (28.0±3.0% vs 26.2±1.7%, P =0.027) were higher in the recurrence group, consistent with more advanced diffuse myocardial disease and inflammation. LA active strain alone predicted AF recurrence (AUC 0.77, P <0.001); a combined model incorporating LA active strain, T1, T2 and ECV further improved discrimination. Patients meeting all four Youden-derived cut-offs (LA active strain >−8%, T1 >1017.3 ms, ECV >27%, T2 >52.4 ms) had a markedly higher risk of AF recurrence (hazard ratio 17.70, 95% CI 3.89–80.61; log-rank P <0.001). Conclusion In patients with preserved ejection fraction undergoing CA for AF, AF recurrence at 1 year was associated with impaired post-ablation LA active function and with LV tissue-mapping abnormalities reflecting diffuse myocardial disease and inflammation. CMR-derived indices of atrial function together with LV native T1, T2 and ECV identify a high-risk post-ablation phenotype and may support substrate-oriented risk stratification after AF ablation.